Since the advent of antiretroviral therapy, a substantial number of HIV+ adults enjoy stable health and an extended lifespan. However, despite the financial and social benefits of employment, studies show that 45 to 62% are unemployed, not all by choice. Barriers are both structural (e.g. the golden handcuffs of needs-based benefits) and personal. In view of the widespread prevalence of fatigue in the HIV population, we conducted clinical trials to treat fatigue with modafinil or armodafinil for 7 years However, we observed that alleviation of fatigue by itself did not enable most patients to find work on their own despite their stated intention to do so. We are now completing a pilot feasibility (R 34) study in which we adapted and refined a stepped program treating fatigue with armodafinil; once fatigue improves, we provide a manualized intervention using the behavioral activation model (Behavioral Activation Program for Energy and Productivity: BA-PEP) to promote achievement of personal goals of employment (paid or volunteer, part-time or more) and/or vocational training. Several patients received supportive counseling to test its acceptability and feasibility as a control arm. Among 44 patients whose energy improved and who received BA-PEP counseling, 63% found work, compared to 28% of patients in earlier fatigue treatment trials of modafinil and armodafinil alone. These findings encourage us to formally assess BA- PEP efficacy. We propose to conduct a randomized clinical trial comparing BA-PEP with supportive counseling, to be conducted in both in our research clinic and a community clinic site, for 100 patients presenting with clinically significant fatigue whose energy has improved with armodafinil, as well as 40 HIV+ patients without significant fatigue at baseline. All participants are HIV+ adults whose health is stable, without current substance use or untreated major depression, and who wish to return to work or vocational training but who have not done so on their own. In addition to testing the efficacy of an intervention easily adapted to other settings with implementation data to support its dissemination to community-based settings, our findings may contribute toward modification of current government policy placing strict limits on earned income to retain needs based benefits.